The FASEB Journal express article 10.1096/fj.00-0866fje. Published online July 9, 2001.

Five-lipoxygenase inhibitors can mediate apoptosis in human breast cancer cell lines through complex eicosanoid interactions Ingalill Avis*, Sung H. Hong* Alfredo Martínez,* Terry Moody,* Yung H. Choi,* Jane Trepel,* Rina Das,† Marti Jett,† and James L. Mulshine* *Department of Cell and Cancer Biology, Medicine Branch, Division of Clinical Sciences, National Cancer Institute, 9000 Rockville Pike, Bethesda, MD 20892-1906; †Walter Reed Army Institute of Research, Division of Pathology, Washington DC, 20307-5100 The views of the authors do not purport to reflect the position of the Department of the Army or the Department of Defense (Para.4-3) AR 360 5.03333. Corresponding author: James L. Mulshine M.D., Intervention Section, CCB, MB, DCS, NCI, NIH, Clinical Center, B10, 12N226, 9000 Rockville Pike, Bethesda, MD 20892-1906. E-mail: mulshinj@mail.nih.gov ABSTRACT Many arachidonic acid metabolites function in growth signaling for epithelial cells, and we previously reported the expression of the major arachidonic acid enzymes in human breast cancer cell lines. To evaluate the role of the 5-lipoxygenase (5-LO) pathway on breast cancer growth regulation, we exposed cells to insulinlike growth factor-1 or transferrin, which increased the levels of the 5-LO metabolite, 5(S)-hydrooxyeicosa-6E,8C,11Z,14Z-tetraenoic acid (5-HETE), by radioimmunoassay and high-performance liquid chromatography. Addition of 5-HETE to breast cancer cells resulted in growth stimulation, whereas selective biochemical inhibitors of 5-LO reduced the levels of 5-HETE and related metabolites. Application of 5-LO or 5-LO activating protein-directed inhibitors, but not a cyclooxygenase inhibitor, reduced growth, increased apoptosis, down-regulated bcl-2, up-regulated bax, and increased G1 arrest. Exposure of breast cancer cells to a 5-LO inhibitor up-regulated peroxisome proliferator-activated receptor (PPAR)α and PPARγ expression, and these same cells were growth inhibited when exposed to relevant PPAR agonists. These results suggest that disruption of the 5-LO signaling pathway mediates growth arrest and apoptosis in breast cancer cells. Additional experiments suggest that this involves the interplay of several factors, including the loss of growth stimulation by 5-LO products, the induction of PPARγ, and the potential activation of PPARγ by interactions with shunted endoperoxides. Key words: breast cancer • leukotrienes • 5-lipoxygenase pathway • peroxisome proliferator-activated receptor • apoptosis

he lifetime risk of breast cancer in American women is higher than for any other malignancy (1). A variety of metabolic and hormonal factors, including dietary fat, are postulated to have a promotional effect on the progression of breast cancer, but how these

factors contribute to the pathogenesis of the disease process is not understood (2–4). Growth factors can function as survival factors and have been reported to inhibit apoptosis (5–7). Insulinlike growth factor-1 (IGF-1) is an important growth factor for breast cancer. Activation of the IGF-type 1 receptor (IGF-R), possibly through the action of phosphatidylinositol 3-kinase has been suggested to be a critical tumor promotion and survival factor (8–13). We previously reported that overexpression of IGF-R and its ligand are conserved features of both breast and lung cancer (14) and that IGF-1-dependent growth stimulation and survival effect can be neutralized by blocking the 5-lipoxygenase (5-LO) pathway of arachidonic acid (AA) metabolism in lung cancer (15). The AA metabolizing enzymes are emerging as significant mediators of growth stimulation for epithelial cells. Earashi and Noguchi suggested that AA metabolism may play a significant role in mammary carcinogenesis through oxidative processes (16, 17), and Przyipiak and co-workers evaluated the effects of 5-LO on the proliferation of MCF-7 cells (18). AA can be metabolized either by the cyclooxygenase (COX) or the LO pathways, and knowledge about the enzymes responsible for both metabolic routes is rapidly increasing (19–21). As part of a general epithelial survey, we recently established the relative frequency of expression of five AA metabolizing enzymes and 5-LO activating protein (FLAP) in breast cancer cells (22). Biologically active products of the 5-LO pathway include 5(S)-hydrooxyeicosa-6E,8C,11Z,14Z-tetraenoic acid (5-HETE) and leukotrienes, which contribute to the inflammatory process in a variety of diseases. Several pharmacological antagonists for the AA pathways that act by different mechanisms are available. The regulation of 5-LO products can be achieved either by direct inhibition of the enzyme such as with the competitive inhibitor AA 861, Zileuton, or by the phenol redox inhibitor, nordihydroguaiaretic acid (NDGA). In addition, another class of selective 5-LO inhibitors exists: MK 886 and MK 591. These are thought to inhibit indirectly by interacting with FLAP and interfering with the presentation of AA to the 5-LO enzyme at the nuclear envelope membrane (20, 23). In light of the importance of the IGF-R pathway in breast cancer and our ability to modulate the signaling for this pathway in lung cancer (15), this study was undertaken to explore the role of the 5-LO pathway in breast cancer cell growth. Our goal is to determine whether targeting the AA oxidation pathway to control breast cancer growth is a viable pharmaceutical strategy. MATERIALS AND METHODS Reagents Synthetic 5-HETE, 5-LO inhibitors NDGA and AA861, and the FLAP inhibitor MK 886 were obtained from BIOMOL Research Laboratories (Plymouth Meeting, PA). The novel FLAP antagonist MK 591 was a kind gift from Merck Frosst Centre for Therapeutic Research (Pointe Claire-Dorval, Quebec, Canada), and Zileuton was kindly provided by the NCI Chemoprevention Drug Repository (Rockville, MD). The COX inhibitor acetylsalicylic acid

T

(ASA) as well as peroxisome proliferator-activated receptor (PPAR) ligands clofibrate and fenofibrate were purchased from Sigma Chemicals (St. Louis, MO). PPAR ligands LY 171883 and WY-14643 were purchased from BIOMOL. Cell lines Cell lines used in the study were obtained from the American Type Culture Collection (Rockville, MD). They included MCF-7, ZR-75, T47 D, SKBR-3, and MB-231. The cells were maintained in RPMI-1640, or MEM Zinc option medium, supplemented with 5% fetal bovine serum (FBS), penicillin (50 units/ml), and streptomycin (50 µg/ml) (Life Technologies, Gaithersburg, MD), in a humidified atmosphere of 95% air and 5% CO2 at 37°C. Growth studies We used a modification of a semiautomated MTT colorimetric assay (Promega, Madison, WI). This assay is based on the reduction of a tetrazolium compound by cells to a colored formazan end product, which is quantitated by measuring the change in optical density (570 nm) compared with a control. Seeding densities were 1–2 × 104 cells/well, and cells were grown for 5 days. Experimental conditions were as previously described (15), with all experiments repeated at least three times. Radioimmunoassay (RIA) analysis for 5(S)-HETE For this assay, the samples were prepared and extracted as previously described (15, 24). In brief, after exposure to the growth factors transferrin (TF) and IGF-1 (10 µg/ml and 5 µg/ml, respectively), the cells were disrupted and extracted onto C-18 disposable cartridges. After washing with 2% ethanol, the eicosanoids were eluted using 85% acetonitrile/15% MeOH. The samples were freeze-dried, reconstituted in ethanol, and diluted in the RIA buffer provided with the RIA kit (Perseptive, Boston, MA). The metabolite was analyzed in duplicate according to the manufacturer’s protocol, and the values obtained were averaged. The RIA antibody does not distinguish 5-HETE from its delta-lactone and methyl-ester derivatives. The experiments were repeated three times. High-performance liquid chromatography (HPLC) characterization These studies were performed as previously described in detail (15, 24). In brief, MCF-7 cells were incubated overnight in medium containing [3H] AA (100 µCi/flask). The culture fluid was removed, and the cells were washed twice with phosphate-buffered saline (PBS). Balanced salt buffer was added to the flask (with and without agonist), and the reaction was stopped after 90 seconds by the addition of 20 µl each of 22 M formic acid and 10 mM butylated hydroxytoluene in methanol. An internal standard, [14C] eicosatrienoic acid, was added to each sample, and the sample was extracted onto disposable C-18 cartridges as previously described (24). The various metabolites were separated by RP-HPLC (Beckman Instruments, Irvine, CA) and quantitated using an in-line radioactivity detector (Packard Radiomatic, Chicago, IL). When inhibitors were used, they were added 30 min before the addition of agonist.

In vitro apoptosis analysis The cells were exposed in vitro to enzyme inhibitors in eight-well cytochambers at cell and drug concentrations comparable to those used in the growth assays. The cells were washed in PBS, fixed in 95% ethanol, and stored frozen until analyzed. In some experiments, cytospins were used instead of cytochambers. The cell lines were analyzed for the presence of apoptosis by using the TUNEL ApopTag in situ apoptosis detection kit (Oncor, Gaithersburg, MD), following the manufacturer’s instructions. The number of apoptotic cells was counted in sets of 100 cells for each experimental well, and the percentage of apoptotic cells was determined. Alternatively, the presence of apoptosis was analyzed using 4,6-diamidino-2-phenylindole (DAPI) staining. The cells were fixed with 3.7% paraformaldehyde in PBS for 10 min at room temperature. Fixed cells were washed twice with PBS, and DAPI (Sigma Chemicals) was added (final concentration 1 µg/ml) for 10 min at room temperature. The cells were washed two more times with PBS and were analyzed via fluorescence microscopy. In parallel, immunocytochemical analysis using antibodies for Bcl-2 and Ba (Santa Cruz Biotechnology, Santa Cruz, CA), two proteins whose expression changes during apoptosis, was performed using a Zymed Histostain kit (Zymed Laboratories, San Francisco, CA). Western blot analysis MCF-7 cells were grown as described for the growth studies and incubated with 4 µM NDGA or 2 µM MK 886 for different time periods. Cells were washed in PBS and lysed. Equal amounts of protein were subjected to electrophoresis on 12% polyacrylamide gels and transferred to nitrocellulose membranes (Schleicher & Schuell, Keene, NH) by electroblotting. Blots were blocked with TNN buffer containing 5% nonfat dried milk and probed with anti-Bcl-2 and anti-Bax antibodies. After washing, the blots were incubated in a horseradish peroxidase-conjugated secondary antibody (Amersham, Arlington Heights, IL). Specific protein bands were visualized with an enhanced chemiluminescence detection system (Amersham, Arlington Heights, IL). Cell cycle analysis MB-231 and MCF-7 cells were treated with inhibitors for 24 h. Cells were harvested and fixed in 70% ethanol at 4°C for 30 min. For FACS analysis, the fixed cells were incubated with 1 ml of PBS containing 50 µg/ml each of RNase A (Sigma Chemicals) and the DNA intercalating dye propidium iodide (Sigma Chemicals) at 4°C for 1 h. The stained cells were analyzed by flow cytometry (FACSCaliber, Becton Dickinson, NC). The percentages of cells in G1, S, and G2-M phases were determined using the ModFit program. The experiment was repeated twice. In vivo apoptosis analysis MCF-7 tumor cells (10

7/mouse) were subcutaneously injected into the flanks of athymic nu/nu

Balb/c mice, and a palpable mass formed after 7 days. Treatment began on day 7 and consisted of three groups of five mice each. One group of five mice was given MK 591, initial dose 2.5 mg/day s.c., for the first week. This dose was reduced threefold for the remainder of the study due to skin toxicity. A second group of mice was given 0.1% NDGA in their drinking water ad

libitum; the third group (placebo group) received 0.1 cc of PBS (s.c.) daily. The duration of the study was 4 weeks, and the tumors were measured biweekly. For apoptosis quantitation, tumors from each of the groups were harvested, fixed in 10% buffered formalin for 24 h, and embedded in paraffin. Tissue sections were analyzed for the presence of apoptosis by using the ApopTag in situ apoptosis detection kit. The number of apoptotic cells was counted in 10 microscopic fields (40× magnification) of each case. Polymerase chain reaction (PCR)-based mRNA analyses for PPAR Cells were grown for 24 or 48 h in the presence or absence of inhibitors at 5 µM concentration. RNA was isolated using the Trizol method (Life Technologies, Gaithersburg, MD), and reverse transcriptase (RT)-PCR was performed using specific primers for PPAR and actin genes. Forward (F) and reverse (R) primers used to detect PPAR cDNAs were PPARα-F (5'-GGCCTCAGGCTATCATTAC-3'), PPARα-R (5'-CCATTTCCATACGCTACC-3'), PPARγ-F (5'-TTCAAACACATCACCCCCC-3'), and PPARγ-R (5'-TTGCCAAGTCGCTGTCATC-3'). The ethidium bromide stained image was digitized and the optical density calculated using the NIH Image program. Values were normalized with the actin value obtained with commercially available primers (Clontech, Palo Alto, CA). These experiments were repeated three times. Northern blot analysis After treatment for the indicated time points, cells were washed with PBS and total RNA was extracted using the RNeasy Mini Kit (QUIAGEN, Valencia, CA). Ten µg of RNA were loaded per lane, run in 1% agarose gels containing 2.2 M formaldehyde, blotted by capillarity onto nitrocellulose membranes (Schleicher & Schuell, Keene, NH), and baked for 2 h at 80°C. Equal loading and integrity of RNA were monitored by ethidium bromide staining. The human PPARγ cDNA probe (Cayman Chemicals, Ann Arbor, MI) was labeled with [α-32P]dCTP (3000 Ci/mmol; NEN Life Science Products, Boston, MA) by random priming. Unincorporated label was removed by Probe Quant G-50 Micro Columns (Amersham Pharmacia Biotech, Piscataway, NJ). Hybridization was carried out overnight at 42°C in Hybrisol 1 (Intergen, Purchase, NY). After stringency washes, blots were exposed to XAR film. Statistical evaluation Significance of difference between samples was determined using Student’s t test. P < 0.05 was regarded as significant. RESULTS Production of 5-HETE in response to growth factors 5-LO rapidly metabolizes AA to hydroperoxy-eicosatetraenoic acid, which is subsequently metabolized to 5-HETE and its derivatives. To demonstrate the presence of bioactive 5-LO enzyme in breast tumor cells, we utilized a specific RIA for 5-HETE. In each of four

representative breast cancer cell lines, stimulation with IGF-1 or TF increased the production of 5-HETE two- to fourfold above control levels (Fig. 1). In three of the four cell lines, the IGF-1-induced production of 5-HETE was greater than the levels of 5-HETE induced by TF. Effect of 5(S)-HETE on breast tumor cell growth To assess the relevance of 5-LO activity on breast cancer, we evaluated the growth effect of exogenous exposure to 5(S)-HETE on breast tumor cells under serum-free conditions. The data summarized in Table 1 demonstrate that the exogenous addition of 5(S)-HETE increased tumor cell proliferation 25–45% above control in four out of five cell lines tested. Other derivatives of the 5-LO pathway such as 5(R)-HETE, 5(±)-HETE, or 5oxoETE had no significant proliferative effect on the cell lines at the concentrations we tested. The most responsive cell line, ZR-75, was also evaluated for growth in response to exposure to leukotriene (LT)D4, and concentrations of 1 nM and 10 nM increased cell growth 25% over control. In contrast, no significant growth stimulatory effects were observed with other AA metabolites, including LTB4, 12(S)-HETE, or 15(S)-HETE under the same experimental conditions (data not shown). Biochemical inhibition of AA metabolism The activity of 5-LO can be blocked in a variety of ways by using defined biochemical inhibitors. Because the FLAP-directed inhibitors are consistently potent and have a more specific mechanism of action, we evaluated the biochemical effect of MK 886 in detail. For these experiments, the AA metabolites generated by IGF-1 stimulation of the cell line MCF-7 were evaluated with or without pre-exposure to MK 886 (Fig. 2). In the absence of the inhibitor (Fig. 2, open bars), those metabolites derived from 5-LO activity were increased 5- to 30-fold, relative to control values after exposure to IGF-1. Oxidation products from other LOs, such as 12- and 15-HETEs and their metabolites, were elevated 5- to 12-fold compared with control. In these experiments, the COX products (thromboxanes and prostaglandins) were also elevated approximately 12- and 21-fold, respectively. This pattern changed upon exposure to MK 886. As shown in Figure 2 (solid bars), MK 886 exposure resulted in shunting of endoperoxide metabolites to alternate pathways as well as stabilization of the parent 5-LO metabolite, 5(S)-hydroperoxyoxyeicosa-6E,8C,11Z,14Z-tetraenoic acid (5-HPETE), relative to cells exposed to IGF-1 alone. However, the 5-LO downstream oxidation products (including 5-HETE and its metabolites) were reduced to below baseline levels. In addition, after exposure to MK 886, the potent LTD4 oxidation products were reduced from 30-fold elevation over controls to about 3-fold. LTD4 oxidation products were also reduced from 6-fold elevation to below control levels. It is of interest to note that for both LTC4 and LTB4 as with 5-HPETE, the first metabolite in each of these pathways is modestly increased after treatment with MK 886, which we attribute to stabilization of the transient species. In contrast, 15-LO products were elevated from 6- to 13-fold (15-HPETE), from 11- to 24-fold (15-HETE and its metabolites), and from 5- to 8-fold (lipoxin A4, a tri HETE). In addition, production of prostanoids was elevated from 21- to 34-fold by the inhibitors, indicating a diversion from the 5-LO pathway to the other metabolic pathways. Effect of 5-LO inhibitors on tumor cell growth

To determine the biological effect of AA metabolic pathway inhibitors on breast cancer cell line growth, we analyzed the effect of several different inhibitors by using a proliferation assay. Table 2 summarizes the average percent growth inhibition for each compound evaluated. We observed significant growth inhibition compared with vehicle control with both FLAP inhibitors. The specificity of this effect was suggested by 5-HETE add-back growth assays in which the growth inhibitor effect of the FLAP inhibitors could be neutralized if 5-HETE was added back within 12 h of drug exposure (Fig. 3). Our analyses also revealed a significant, potent, and reproducible growth inhibition with 5 µM NDGA on all cell lines tested. With the direct 5-LO enzymatic inhibitors, AA861 and Zileutin, we observed more modest growth inhibition in two of five and three of four cell lines, respectively (Table 2). The drug concentration (10 µM) required for these growth inhibitors was higher than with the other LO inhibitors. For comparative analysis, the contribution of the COX pathway for breast cancer was evaluated in the presence of ASA. As summarized in Table 2, the breast cancer cell lines were not significantly growth inhibited even by the addition of high doses (100 µM) of ASA. Evidence of apoptosis in vitro To test the hypothesis that an apoptotic pathway was involved in response to the LO-inhibitors, we examined breast cancer cell lines in vitro after exposure to two inhibitors. Figure 4 summarizes the percentage of apoptotic cells determined in five breast cancer cell lines. An increase in the number of apoptotic bodies was observed in all cell lines tested with both inhibitors. Figure 5 demonstrates the cytomorphological changes induced by treatment with MK 886 and NDGA on cell lines MCF-7 (Fig. 5, A–C) and MB-231 (Fig. 5, D–F). The cells were stained with the intercalating stain DAPI (Fig. 5, A–F). The addition of AA metabolism inhibitors resulted in changes consistent with the appearance of apoptotic bodies. Figure 5 (G–I), demonstrates the staining pattern obtained by application of the TUNEL assay after exposure to MK 886 and NDGA, confirming the presence of apoptotic cells after exposure to the inhibitors. Similar morphologic features were observed after treatment with a second FLAP inhibitor, MK 591 (data not shown). To further examine the role of LO inhibitors in inducing apoptosis, we evaluated the expression of bcl-2 in MCF-7 cells as shown in Figure 6. These cells have been reported to express high levels of bcl-2 (25). Using immunocytochemical analysis, we observed down-regulation of bcl-2 in MCF-7 cells treated with a FLAP inhibitor (MK 886) compared with untreated cells in which bcl-2 expression is clearly observed (Fig. 6A, B). In addition, bax expression was induced by exposure of the breast cancer cells to MK 886 compared with untreated cells (Fig. 6C, D). To confirm this observation, we examined the levels of bcl-2 and bax by Western blot analysis after treatment of the cell line MCF-7 with NDGA and a FLAP inhibitor. As shown in Figure 6E, treatment of MCF-7 with the inhibitors resulted in a decrease of bcl-2 and a concomitant increase in bax immunoreactivity. These results were comparable whether either of the FLAP inhibitors, MK 886 or MK 591, is used. Studying the cell cycle status of MB-231 and MCF-7 cells, as shown in Table 3, provided additional evaluation of the mechanism of apoptosis. For the cell line MB-231, treatment with both 4 µM NDGA and 2 µM MK 886 resulted in an increase in cells found in G1 compared with

untreated cells. There was also a statistically significant decrease in the number of cells found in S and G2/M phase for these dose concentrations. In cell line MCF-7, treatment with the two inhibitors had less impact, and the observed changes did not reach statistical significance. Evidence for induction of apoptosis and tumor reduction in vivo To confirm the role of LO inhibitors as inducers of apoptosis in vitro, we utilized an in vivo model system. Athymic nu/nu mice bearing heterotransplants of the breast cancer cell line MCF-7 were administered the LO inhibitor NDGA, the FLAP inhibitor MK 591, or a placebo. The tumor xenografts were harvested and examined for the presence of apoptosis-induced oligonucleosomes. Figure 7A shows that in both groups of mice treated with either NDGA or MK 591, there was a significant difference in the number of cells with apoptotic nuclei between the treated and the placebo group (P<0.05). This finding correlated with a statistically significant reduction in tumor size in the group of mice that were given NDGA (Fig. 7B). The flank tumor size was an average of 765 ± 142 mm3 for the NDGA treated group and was 2394 ± 591mm3 for the placebo group (P<0.05). In the mice receiving MK 591, there was close to a 50% reduction in tumor size compared with placebo, but this difference did not achieve statistical significance (P>0.05). However, the dose of MK 591 had to be reduced by 66% after the first week of drug administration because of unexpected cutaneous ulceration at the injection site. This skin condition resolved spontaneously with dose reduction. PPAR induction In light of recent reports regarding the mechanistic basis of the antiproliferative effect of the FLAP inhibitor (26), we explored alternative mechanisms for the growth effects on 5-LO inhibition. We noted a large increase in 15-HETE production in response to the exposure to MK 886 (Fig. 2), and 15-HETE has been proposed to be a ligand for PPARγ (27). When we exposed the breast tumor cell line ZR-75 to 5 µMK 886 or NDGA for 24 and 48 h, an up-regulation of both PPARα and PPARγ expression occurred. The biggest increase could be observed after a 48-h exposure with both inhibitors (Fig. 8A–D). Effect of PPAR ligands on breast tumor cell growth To further evaluate the possible involvement of PPAR in the growth regulation of breast cancer cell lines, we tested a range of selective PPAR agonists for their effect on breast cancer cell lines. This panel included ligands for PPARα (WY-14643, clofibrate, fenofibrate) and PPARγ (LY 171883). When breast cancer cell lines T47D and ZR-75 were incubated with each of the four PPAR ligands, a dose-dependent growth reduction was observed with all the compounds for both cell lines compared with vehicle control (Fig. 8E, F). At the higher doses, growth inhibition ranging from 60 to 80% were observed. Interaction of PPAR effects on breast cancer cell growth Because the induction of PPARs occurs promptly with exposure to 5-LO inhibitors, we explored further to determine whether that could be involved in the breast cancer cell growth regulation. From the Northern blot analysis (Fig. 9A), induction of PPARγ is evident within 6 h of exposure

of T47D cells to the most potent 5-LO inhibitor, MK886. This finding is confirmed by semiquantitative RT-PCR for PPARγ (Fig. 9B), but in these experiments, the induction of PPARγ is more protracted. We next asked whether the up-regulation of PPARγ could be contributing to the growth inhibition of the T47D cells. In Figure 9C, the filled circles represent cells exposed sequentially for 12 h to 5 µM MK886 and then for 24 h to 4 µM LY 171883, PPARγ ligand. The sequential exposure to the FLAP inhibitor followed by the PPARγ ligand is associated with significantly more growth inhibition than exposure to the ligand alone or the FLAP inhibitor alone for the same amount of time (Fig. 9C). Under the same experimental conditions, we evaluated the impact of the sequential exposure of these drugs on apoptosis (Fig. 9D). The sequential drug exposure over 36 h was more potent than comparable exposure to MK-886 alone. Over this same time course, exposure of the breast cancer cells to 4 µM LY 171883 had no effect on increasing apoptosis (data not shown). These experiments suggest that the consequences of endoperoxide shunting (Fig. 2) can generate products capable of interacting with the upregulated PPARγ and have significant growth effects potentially through enhanced apoptosis. DISCUSSION We are interested in mediators that function as survival factors, because this may be a central mechanism of carcinogenesis and therefore constitute a critical target for effective chemoprevention. Exogenously added IGF-1 and TF increased the endogenous production of 5-HETE and its derivatives in breast tumor cell lines. The exogenous addition of 5(S)-HETE and the cysteinyl leukotrienes was consistently growth stimulatory to breast cancer cells cultured under serum-free conditions. The extent of growth stimulatory effects was modest (20–45%) but comparable to a magnitude of growth stimulation with other important cancer cell growth mediators (28). Previously, we demonstrated that a panel of breast cancer cell lines, including all of the lines used in this report, express the message for both 5-LO and FLAP (22). Collectively, these results support the hypothesis that a product of the 5-LO enzyme can stimulate breast cancer cell growth. HPLC analysis confirmed that exposure of breast cancer cells to IGF-1 increases the production of the relevant AA metabolites in MCF-7 cell cultures. The greatest increases were seen for the major 5-LO metabolic pathways, resulting in production of 5-HETE, LTC4, and their metabolites. The individual effects of specific metabolites have been documented in this and other studies. LXA4, a tri-hydroxyETE formed by joint activity of 5- and 15-LO, has previously been shown to interact with specific G protein-coupled receptors on cell surfaces (29). The finding of increased levels of LTC4 is significant because this metabolite is an activator of protein kinase C γ (PKCγ) (30) and PKCγ is involved in the regulation of cell proliferation (31–33). These findings taken collectively support the notion that 5-LO products released in response to growth factors may regulate cellular proliferation in cancer. In response to IGF-1 exposure in the presence of a FLAP inhibitor, MK 886, eicosanoid metabolism was significantly altered. Whereas parent compounds (LTC4 and LTB4) in the 5-LO pathway accumulated under these conditions, their subsequent metabolic products were at baseline levels and significantly decreased (about 7- to 10-fold). These findings may have several interpretations. For example, the interplay between the various metabolites may account for some of the effects observed in response to IGF-1. A previous study suggests structural similarity between FLAP and LTC4 synthase and that the FLAP inhibitor MK 886 can inhibit

LTC4 synthase (34). In addition, the elevated concentration of the parent compounds may be attributable to accumulation due to the inhibition of their usual metabolizing enzymes, or by stabilization of these transient compounds by the conjugated ring structures of the drug. Potentially more important, the downstream 5-LO metabolites were greatly reduced due to diversion of the metabolic products from 5-LO to other LO (12-LO and 15-LO metabolites) and COX pathways, similar to previous reports in other cell systems (15, 24). Many complex interactions may be occurring, which are especially relevant with the endoperoxide shunting of substrate to other AA metabolic pathways (35). We used biochemical inhibitors in this study, because these would be the most obvious tools for clinical evaluation of strategies directed at modulating the 5-LO pathway. Molecular approaches to study gene families, including many closely conserved gene family members, have been confounded by the complexity of this system (36, 37). An unresolved issue in this regard is how to distinguish the effect of molecularly engineered loss of specific enzyme function from the secondary changes in substrate availability or utilization (37). Most of what is known about the effects of AA metabolites relates to their functions in inflammation rather than in cancer systems, but this inflammatory biology may also be relevant to carcinogenesis (38–41). In this study, the consistent in vitro consequence of inhibition of 5-LO was increased apoptosis and corresponding growth inhibition. In contrast, the COX inhibitor, ASA, did not inhibit breast cancer cell line growth, even at high concentrations of up to 100 µM. This is consistent with previous reports on other epithelial systems (42–45). We recently reported a discrepancy with COX inhibitors between in vitro and in vivo effects in oropharyngeal cancer. This finding suggests that COX inhibitors function through an interaction between epithelial and immune cells that is not reflected in simple in vitro assays (45); therefore, the current report does not preclude an important contribution of COX activity to breast carcinogenesis. A valid concern about this work is the question about the specificity of drug effect with all of the 5-LO inhibitors but in particular with NDGA. NDGA has previously been shown to inhibit carcinogenesis in other systems (46, 47), but NDGA also disrupts a number of iron-dependent AA metabolic pathways (48). To address this specificity concern, we evaluated several biochemical inhibitors that act through different mechanisms. The FLAP inhibitors MK 886 and MK 591 were potent inhibitors (<5 µM concentrations of the drug). These inhibitors have been reported to interfere with the presentation of 5-LO at the nuclear envelope. Our data demonstrate that these drugs effectively induced cell cycle arrest in G1 and apoptosis in breast cancer cells consistent with a previous report involving megakaryoblastic leukemia (49). Because all the different types of 5-LO inhibitors affected growth inhibition and apoptosis, this result supports involvement of 5-LO in this biology. Multiple mechanisms can induce apoptosis (50), including modulation of the LO pathways in a rat Walker (W256) carcinosarcoma cells system (51). Inhibitors of 5-LO reactions may act through several mechanisms, including redox interactions (48). This information suggested the importance of evaluating the bcl-2/bax relationship, because these genes are involved with regulation of redox pathways (52). Some proto oncogenes (e.g., bcl-2 and bcl-xL) of the bcl-2 family suppress apoptosis induced by various oxidative processes (53, 54). The gene for bcl-2 is overexpressed in 70% of breast cancers and has been postulated to play a role in breast cancer

development (55). We observed, using immunocytochemistry and Western blot analysis, that bcl-2 expression is down-regulated after exposure to LO inhibitors. These findings correlated with a reduction of tumor growth. Induction of differentiation and apoptosis in cancer cells can also occur through the action of other oxidation products of AA. PPARα and PPARγ are novel nuclear hormone receptors activated by long-chain fatty acids and synthetic ligands, which regulate lipid metabolism and have been shown to be expressed in breast cancer cells (56). Mueller et al. showed that ligand activation of PPARγ in cultured breast cancer cells causes extensive lipid accumulation; changes in breast epithelial gene expression associated with a more differentiated, less malignant state; and a reduction in growth rate and clonogenic capacity of the cells (56). In our experiments, doses of LO inhibitors that result in apoptosis and growth inhibition also induce regulation of both PPARα and PPARγ. To evaluate the validity of this response, we examined the effect of specific PPAR agonists on relevant breast cancer cell lines and observed dose-dependent growth inhibition similar to that observed for the 5-LO inhibitors. These data suggest that interrupting the 5-LO pathway may also activate the PPAR transcriptional pathway. Activation of PPARγ through a synthetic ligand, troglitazone, and other PPARγ activators causes inhibition of proliferation and apoptosis both in vitro and in vivo, concomitant with a reduction of bcl-2 protein expression (57). The importance of PPAR activation, including PPARδ, contributing to epithelial carcinogenesis is evident in other tumor systems (58–61). To further explore alternative mechanisms for the growth effects of pharmacological inhibition of 5-LO activity, we conducted additional experiments to explore the impact on the PPARγ pathway at time points before evidence of increased apoptosis. After exposure of the breast tumor cell line T47D to 5 µm MK 886, PPARγ was up-regulated by 6 h, as revealed by both Northern and PCR. A range of selective PPAR agonists demonstrated growth inhibition in 4-day assays. This panel included ligands for PPARα (WY-14643, clofibrate, fenofibrate) and PPARγ (LY 171883). To evaluate if the PPARγ up-regulation was contributing to the earlier growth inhibitory effect of the 5-LO inhibitors, we exposed T47D cells for 12 h to MK 886. After washing the cells, we incubated them with the PPARγ ligand, LY171883. This sequential drug exposure was significantly more antiproliferative and apoptotic then were the parallel exposures to either FLAP inhibitor or PPARγ ligand alone. Therefore, the enhanced inhibitory effect of sequential drug exposure may be due to the interaction of the PPARγ induction occurring in the face of increased production by endoperoxide shunting of alternative eicosanoid products capable of activating PPARγ. Recently, there was a suggestion that drugs, such as MK 886, did not mediate their antiproliferative effects through FLAP (26). The basis for this concern was due to the high doses of FLAP required for in vivo effects and the fact that a cell line without FLAP expression was none the less growth inhibited by these drugs. The case against the involvement of FLAP in mediating apoptosis may not be universally applicable, because there is a large body of data to the contrary. Biological studies in mice engineered to be FLAP deficient support a critical role for FLAP in the production of leukotrienes (41). Although much higher drug doses with the FLAP inhibitors are required in vivo than would be expected from the effective in vitro dose, this same situation also occurs with celecoxib and is thought to relate to complex pharmacology and serum protein interactions rather than mechanistic infidelity (62, 63). Recent reports expand the

role of FLAP in the utilization of 12-HETE and 15-HETE (64) and in COX-2 regulation (65). In addition, other mechanisms for FLAP action have been recently proposed, including inhibition of LTC4 synthase activity and K+ channels (66, 67). This situation becomes even more complex in considering the implications of the PPAR activation. We have already noted the complexity of alternative substrate utilization by AA metabolizing enzymes. A range of AA substrates at critical concentrations could function as PPAR ligands, as suggested by Huang and co-workers (27). The relatively nonselective nature of the PPAR receptors and the permissive requirement to activate transcriptional activation of PPAR response elements (68–70) present a new challenge in considering this biology as a target for growth regulation. Given the myriad effects of PPARs (71–73) and the difficulty in dissecting distinct fatty acid effects (37), sorting out these issues with confidence will take considerable further study. In summary, interference with 5-LO activity is associated with increased apoptosis and tumor growth inhibition in breast cancer cells. This may involve the interplay of loss of growth stimulation by 5-LO products and the potential recruitment of the inhibitory effect of PPARs, especially PPARγ. Further study allowing more precise delineation of the individual contribution of the AA metabolites is required, but the implications of extensive cross-talk and cross-activation as inherent aspects of this lipid biochemistry need to be considered in contemplating any future “specific” therapeutic intervention. Although these important and complicated issues are being addressed, we will focus on the impact of biochemical inhibitors on functional end points, such as modulation of apoptosis as the critical discriminant in guiding translational research in this area. ACKNOWLEDGMENTS We thank Dr. R. Murphy for helpful discussions. We also thank Merck Frosst Centre for Therapeutic Research (Pointe Claire-Dorval, Quebec, Canada), for the kind gift of FLAP inhibitor MK 591. In addition, we are grateful to Amy Guzzone and Valerie McKey for excellent technical assistance. REFERENCES 1. Menick, H., Jessup, J., Eyre, H., Cunningham, M., Fremgen, A., Murphy, G., and Winchester,

D. (1997) Clinical highlights from the National Cancer Data Base. CA Cancer J. Clin. 47, 161–170

2. Meade, E., McIntyre, T., Zimmerman, G., and Prescott, S. (1999) Peroxisome proliferators

enhance cycooxygenase-2 expression in epithelial cells. J. Biol. Chem. 274, 8328–8334 3. Yano, T., Pinski, J., Groot, K., and Schally, A. (1992) Stimulation by bombesin and inhibition

by bombesin/gastrin-releasing peptide antagonist RC-3095 of growth of human breast cancer cell lines. Cancer Res. 52, 4545–4547

4. Martin, J., Coverley, J. A., Pattison, S., and Baxter, R. (1995) Insulin-like growth factor-

binding protein-3 production by MCF-7 breast cancer cells: stimulation by retinoic acid and

cyclic adenosine monophosphate and different effects of estradiol. Endocrinology 136, 1219–1226

5. Thompson, C. (1995) Apoptosis in the pathogenesis and treatment of disease. Science 267,

1456–1462 6. Piazza, G., Rahm, A., Krutzsch, M., Sperl, G., Paranka, N., Gross, P., Brendel, K., Burt, R.,

Alberts, D., Pamukcu, R., and Ahnen, D. (1995) Anti-neoplastic drugs Sulindac Sulfide and Sulfone inhibit cell growth by inducing apoptosis. Cancer Res. 55, 3110–3116

7. Wang, T., and Phang, J. (1995) Effects of estrogen on apoptotic pathways in human breast

cancer cell line MCF-7. Cancer Res. 55, 2487–2489 8. Cullen, K., Yee, D., Sly, W., Perdue, J., Hampton, B., Lippman, M., and Rosen, N. (1990)

Insulin-like growth factor receptor expression and function in human breast cancer. Cancer Res. 50, 48–53

9. Yee, D., Rosen, N., Favoni, R., and Cullen, K. (1991) The insulin-like growth factors, their

receptors, and their binding proteins in human breast cancer. Cancer Treat. Res. 53, 93–106 10. Baserga, R. (1994) Oncogenes and the strategy of growth factors. Cell 79, 927–930 11. Yee, D., Paik, S., Lebovic, G., Marcus, R., Favoni, R., Cullen, K., Lippman, M., and Rosen,

N. (1989) Analysis of insulin-like growth factor gene expression in malignancy: evidence for a paracrine role in human breast cancer. Mol. Endocrinol. 3, 509–517

12. Gooch, J. L., Van Den Berg, C. L., and Yee, D. (1999) Insulin-like growth factor (IGF)-I

rescues breast cancer cells from chemotherapy-induced cell death--proliferative and anti-apoptotic effects. Breast Cancer Res. Treat. 56, 1–10

13. Lee, A. V., Gooch, J. L., Oesterreich, S., Guler, R. L., and Yee, D. (2000) Insulin-like growth

factor I-induced degradation of insulin receptor substrate 1 is mediated by the 26S proteasome and blocked by phosphatidylinositol 3'-kinase inhibition. Mol. Cell Biol. 20, 1489–1496

14. Quinn, K., Treston, A., Unsworth, E., Miller, M., Vos, M., Griley, C., Battey, J., Mulshine,

J., and Cuttitta, F. (1996) Insulin-like growth factor expression in human cancer cell lines. J. Biol. Chem. 271, 11477–11483

15. Avis, I. M., Jett, M., Boyle, T., Vos, M. D., Moody, T., Treston, A. M., Martinez, A., and

Mulshine, J. L. (1996) Growth control of lung cancer by interruption of 5-lipoxygenase-mediated growth factor signaling. J. Clin. Invest. 97, 806–813

16. Earashi, M., Noguchi, M., Kinoshita, K., and Tanaka, M. (1995) Effects of eicosanoid

synthesis inhibitors on the in vitro growth and prostaglandin E and Leukotrine B secretion of a human breast cancer cell line. Oncology 52, 150–155

17. Noguchi, M., Rose, D., Earashi, M., and Miyazaki, I. (1995) The role of fatty acids and

eicosanoid synthesis inhibitors in breast carcinoma. Oncology 52, 265–271 18. Przylipiak, A., Hafner, J., Przylipiak, J., Kohn, F. M., Runnebaum, B., and Rabe, T. (1998)

Influence of 5-lipoxygenase on in vitro growth of human mammary carcinoma cell line MCF-7. Gynecol. Obstet. Invest. 46, 61–64

19. Dixon, R., Jones, R., Diehl, R., Bennett, C., Kargman, S., and Rouzer, C. (1988) Cloning of

cDNA for human 5-lipoxygenase. Proc. Natl. Acad. Sci. USA 85, 416–420 20. Ford-Hutchinson, A. (1994) Regulation of leukotriene biosynthesis. Cancer and Metastasis

Reviews 13, 257–267 21. Dubois, R., Abramson, S., Crofford, L., Gupta, R., Simon, L., Putte, L. V. D., and Lipsky, P.

(1998) Cyclooxygenase in biology and disease. FASEB J. 12, 1063–1073 22. Hong, S. H., Avis, I., Vos, M. D., Martinez, A., Treston, A. M., and Mulshine, J. L. (1999)

Relationship of arachidonic acid metabolizing enzyme expression in epithelial cancer cell lines to the growth effect of selective biochemical inhibitors. Cancer Res. 59, 2223–2228

23. Peters-Golden, M. and Brock T.G., (2000) Intracellular compartmentalization of leukotriene

biosynthesis. Am. J. Respir. Crit. Care Med. 161, S36–S40 24. Boyle, T., Lancaster, V., Hunt, R., Gemski, P., and Jett, M. (1994) Method for simultaneous

quantitation of platelet activating factor and multiple arachidonate metabolites from small samples: analysis of effects of Staphylococcus aureus enterotoxin B in mice. Anal. Biochem. 216, 373–382

25. Monaghan, P., Robertson, D., Amos, T., Dyer, M., Mason, D., and Greaves, M. (1992)

Ultrastructural localization of bcl-2 protein. J. Histochem. Cytochem. 40, 1819–1825 26. Datta, K., Biswal, S. S., and Kehrer, J. P. (1999) The 5-lipoxygenase-activating protein

(FLAP) inhibitor, MK886, induces apoptosis independently of FLAP. Biochem. J. 340, 371–375

27. Huang, J. T., Welch, J. S., Ricote, M., Binder, C. J., Willson, T. M., Kelly, C., Witztum, J.

L., Funk, C. D., Conrad, D., and Glass, C. K. (1999) Interleukin-4-dependent production of PPAR-gamma ligands in macrophages by 12/15-lipoxygenase. Nature 400, 378–382

28. Nakanishi, Y., Cuttitta, F., Kasprzyk, P., Treston, A., Avis, I., Minna, J., Kleinman, H., and

Mulshine, J. (1988) The effects of growth factors on the in vitro growth of small cell lung cancer as determined in a colorometric assay. In Biology of Lung Cancer (Rosen, S., Mulshine, J., and Abrams, P., eds) Vol. 37, pp. 59–89, Marcel Dekker, New York, NY

29. Serhan, C. (1994) Lipoxin biosynthesis and its impact in inflammatory and vascular events. Biochim. Biophys. Acta 1212, 1–25

30. Shearman, M., Naor, Z., Sekiguchi, K., Kishimoto, A., and Nishizuka, Y. (1989) Selective

activation of the g-subspecies of protein kinase C from bovine cerebellum by arachidonic acid and its lipoxygenase metabolites. FEBS Lett. 243, 177–182

31. Weinstein, I., Bonner, C., Krauss, R., and al, e. (1991) Pleiotropic effects of protein kinase C

and the concept of carcinogenesis as a progressive disorder in signal transduction. In Origins of human cancer: A comprehensive review (Brugge, J., Curran, T., Harrow, E., and Al, E., eds), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY

32. Clemens, M., Trayner, I., and Menaya, J. (1992) The role of protein kinase C isoenzymes in

the regulation of cell proliferation and differentiation. J. Cell Sci. 4, 881–887 33. Newton, A. (1995) Protein kinase C: Structure, function, and regulation. J. Biol. Chem. 270,

28495–28498 34. Lam, B., Penrose, J., Freeman, G., and Austen, K. (1994) Expression cloning of cDNA for

human leukotriene C4 synthase, an integral membrane protein conjugating reduced glutathione to leukotriene A4. Proc. Natl. Acad. Sci. USA 91, 7663–7667

35. Hussey, H., and Tisdale, M. (1996) Inhibition of tumor growth by lipoxygenase inhibitors.

Br. J. Cancer 74, 683–687 36. Brash, A. R. (1999) Lipoxygenases: occurrence, functions, catalysis, and acquisition of

substrate. J. Biol. Chem. 274, 23679–23682 37. Duplus, E., Glorian, M., and Forest, C. (2000) Fatty Acid Regulation of Gene Transcription.

J. Biol. Chem. 275, 30749–30752 38. Cordon-Cardo, C., and Prives, C. (1999) At the crossroads of inflammation and

tumorigenesis [comment]. J. Exp. Med. 190, 1367–1370 39. Hudson, J. D., Shoaibi, M. A., Maestro, R., Carnero, A., Hannon, G. J., and Beach, D. H.

(1999) A proinflammatory cytokine inhibits p53 tumor suppressor activity [see comments]. J. Exp. Med. 190, 1375–1382

40. Hanahan, D., and Weinberg, R. A. (2000) The hallmarks of cancer. Cell 100, 57–70 41. Byrum, R. S., Goulet, J. L., Griffiths, R. J., and Koller, B. H. (1997) Role of the 5-

lipoxygenase-activating protein (FLAP) in murine acute inflammatory responses. J. Exp. Med. 185, 1065–1075

42. Smith, C. J., Morrow, J. D., Roberts, L. J. d., and Marnett, L. J. (1993) Differentiation of monocytoid THP-1 cells with phorbol ester induces expression of prostaglandin endoperoxide synthase-1 (COX-1). Biochem. Biophys. Res. Commun. 192, 787–793

43. Tsujii, M., Kawano, S., Tsuji, S., Sawaoka, H., Hori, M., and DuBois, R. N. (1998)

Cyclooxygenase regulates angiogenesis induced by colon cancer cells [published erratum appears in Cell 1998, 94(2), following 271] Cell 93, 705–716

44. Wallace, J. L., Bak, A., McKnight, W., Asfaha, S., Sharkey, K. A., and MacNaughton, W. K.

(1998) Cyclooxygenase 1 contributes to inflammatory responses in rats and mice: implications for gastrointestinal toxicity [see comments]. Gastroenterology 115, 101–109

45. Hong, S. H., Ondrey, F. G., Avis, I. M., Chen, Z., Loukinova, E., Cavanaugh, P. F., Jr., Van

Waes, C., and Mulshine, J. L. (2000) Cyclooxygenase regulates human oropharyngeal carcinomas via the proinflammatory cytokine IL-6: a general role for inflammation? FASEB J. 14, 1499–1507

46. Ondrey, F., Harris, J., and Anderson, K. (1989) Inhibition of U937 eicosanoid and DNA

synthesis by 5,8,11,14-Eicosatetraynoic acid, an inhibitor of arachidonic acid metabolism and its partial reversal by leukotriene C4. Cancer Res. 49, 1138–1142

47. Yamamoto, S., and Kato, R. (1992) Inhibitors of the arachidonic acid cascade and their

chemoprevention of skin cancer. In Cancer Chemoprevention (Wattenberg, L., Lipkin, M., Boone, C., and Kelloff, G., eds) pp. 141–151, CRC Press Inc, Boca Raton, FL

48. Ford-Hutchinson, A., Gresser, M., and Young, R. (1994) 5-lipoxygenase. Annu. Rev.

Biochem. 63, 383–417 49. Miller, A., Allen, B., and Ziboh, V. (1997) Lipoxygenase metabolism is required for

interlukin-3 dependent proliferation and cell cycle progression of the human M-07e cell line. J. Cell Physiol. 170, 309–315

50. Williams, G., and Smith, C. (1993) Molecular regulation of apoptosis: Genetic controls on

cell death. Cell 74, 777–779 51. Tang, D., Chen, Y., and Honn, K. (1996) Arachidonate lipoxygenases as essential regulators

of cell survival and apoptosis. Proc. Natl. Acad. Sci. USA 93, 5241–5246 52. Korsmeyer, T., Shutter, J., Veis, D., Mery, D., and Oltvai, Z. (1993) Bcl-2/Bax: a rheostat

that regulates an anti-oxidant pathway and cell death. Cancer Biology 4, 327–332 53. Chao, D., Linette, G., Boise, L., White, L., Thompson, C., and Korsmeyer, S. (1995) Bcl-XL

and Bcl-2 repress a common pathway of cell death. J. Exp. Med. 182, 821–828 54. Minn, A., Rudin, C., Boise, L., and Thompson, C. (1995) Expression of bcl-xL can confer a

multidrug resistance phenotype. Blood 86, 1903–1910

55. Sumantran, V., Ealovega, M., Nunez, G., Clarke, M., and Wicha, M. (1995) Overexpression

of Bcl-xs sensitizes MCF-7 cells to chemotherapy-induced apoptosis. Cancer Res. 55, 2507–2510

56. Mueller, E., Sarraf, P., Tontonoz, P., Evans, R., Martin, K., Zhang, M., Fletcher, C., Singer,

S., and Spiegelman, B. (1998) Terminal differentiation of human breast cancer through PPAR gamma. Mol. Cell 3, 464–470

57. Elstner, E., Muller, C., Koshizuka, K., Williamson, E., Park, D., Assou, H., Shintaku, P.,

Said, J., Heber, D., and Koeffler, H. (1998) Ligands for peroxisome proliferator-activated receptor g and retinoic acid receptor inhibit growth and induce apoptosis of human breast cancer cells in vitro and in BNX mice. Proc. Natl. Acad. Sci. USA 95, 8806–8811

58. Brockman, J. A., Gupta, R. A., and Dubois, R. N. (1998) Activation of PPARgamma leads to

inhibition of anchorage-independent growth of human colorectal cancer cells [see comments]. Gastroenterology 115, 1049–1055

59. He, T. C., Chan, T. A., Vogelstein, B., and Kinzler, K. W. (1999) PPARdelta is an APC-

regulated target of nonsteroidal anti-inflammatory drugs. Cell 99, 335–345 60. Inoue, H., Tanabe, T., and Umesono, K. (2000) Feedback control of cyclooxygenase-2

expression through PPARgamma [In Process Citation]. J. Biol. Chem. 275, 28028–28032 61. Chang, T. H., and Szabo, E. (2000) Induction of differentiation and apoptosis by ligands of

peroxisome proliferator-activated receptor gamma in non-small cell lung cancer. Cancer Res. 60, 1129–1138

62. McAdam, B. F., Catella-Lawson, F., Mardini, I. A., Kapoor, S., Lawson, J. A., and

FitzGerald, G. A. (1999) Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: the human pharmacology of a selective inhibitor of COX-2 [published erratum appears in Proc. Natl. Acad. Sci. USA, 1999, 96(10), 5890] Proc. Natl. Acad. Sci. USA 96, 272–277

63. Bishop-Bailey, D. (2000) Peroxisome proliferator-activated receptors in the cardiovascular

system. Br. J. Pharmacol. 129, 823–834 64. Mancini, J. A., Waterman, H., and Riendeau, D. (1998) Cellular oxygenation of 12-

hydroxyeicosatetraenoic acid and 15-hydroxyeicosatetraenoic acid by 5-lipoxygenase is stimulated by 5-lipoxygenase-activating protein. J. Biol. Chem. 273, 32842–32847

65. Battu, S., Beneytout, J. L., Pairet, M., and Rigaud, M. (1998) Cyclooxygenase-2 up-

regulation after FLAP transfection in human adenocarcinoma cell line HT29 cl.19A. FEBS Lett. 437, 49–55

66. Gupta, N., Nicholson, D. W., and Ford-Hutchinson, A. W. (1997) Pharmacological cross-reactivity between 5-lipoxygenase-activating protein, 5-lipoxygenase, and leukotriene C4 synthase. Can. J. Physiol. Pharmacol. 75, 1212–1219

67. Smirnov, S. V., Knock, G. A., and Aaronson, P. I. (1998) Effects of the 5-lipoxygenase

activating protein inhibitor MK886 on voltage-gated and Ca2+-activated K+ currents in rat arterial myocytes. Br. J. Pharmacol. 124, 572–578

68. Lehmann, J. M., McKee, D. D., Watson, M. A., Willson, T. M., Moore, J. T., and Kliewer, S.

A. (1998) The human orphan nuclear receptor PXR is activated by compounds that regulate CYP3A4 gene expression and cause drug interactions. J. Clin. Invest. 102, 1016–1023

69. Kliewer, S. A., Moore, J. T., Wade, L., Staudinger, J. L., Watson, M. A., Jones, S. A.,

McKee, D. D., Oliver, B. B., Willson, T. M., Zetterstrom, R. H., Perlmann, T., and Lehmann, J. M. (1998) An orphan nuclear receptor activated by pregnanes defines a novel steroid signaling pathway. Cell 92, 73–82

70. Kliewer, S. A., Lehmann, J. M., Milburn, M. V., and Willson, T. M. (1999) The PPARs and

PXRs: nuclear xenobiotic receptors that define novel hormone signaling pathways. Recent Prog. Horm. Res. 54, 345–367

71. Wu, G. D., and Lazar, M. A. (1998) A gut check for PPARgamma [editorial; comment].

Gastroenterology 115, 1283–1285 72. Desvergne, B., and Wahli, W. (1999) Peroxisome proliferator-activated receptors: nuclear

control of metabolism. Endocr. Rev. 20, 649–688 73. Corton, J. C., Anderson, S. P., and Stauber, A. (2000) Central role of peroxisome

proliferator-activated receptors in the actions of peroxisome proliferators. Ann. Rev. Pharmacol. Toxicol. 40, 491–518

Received December 12, 2000; revised May 9, 2001.

Table 1 Effect of Exogenous 5-HETE on Growth Stimulation ________________________________________________________________________ Cell Lines 5(S)-HETE (µM) % Maximum Growth ________________________________________________________________________ MCF-7 0.1

140 ± 8*

ZR-75 0.1

145 ± 13*

T47D 0.01

125 ± 1*

SKBR3 0.1

111 ± 3

MB-231 0.02

130 ± 9*

________________________________________________________________________ The second column shows the 5(S)-HETE concentration that produced the maximum growth stimulation. The third column shows percent maximum growth (mean ± s.d.), which was calculated from the optical density value of a minimum of 6 replicates, from at least three differe1t experiments per cell line. *Percent growth proliferation compared to vehicle control (100 % growth) was statistically significant (P<0.05).

Table 2 Growth Inhibition in the Presence of Specific Inhibitors of AA Metabolic Pathways Inhibitors

NDGA AA861 Zileutin MK886 MK 591 ASA 5 µM 10 µM 10 µM 5 µM 5 µM 100 µM Cell Lines MCF-7 74 ± 13* 58 ± 12* NT 88 ± 7* 100 ± 2* <10 ZR-75 57 ± 28* 33 ± 6* 20 ± 5* 92 ± 10* 95 ± 5* <10 T47 D 83 ± 6* <10 20 ± 4* 53 ± 6* 60 ± 2* <10 SKBR3 73 ± 12+ <10 <10 77 ± 15* 90 ± 2* <10 MB-231 65 ±21* <10 20 ± 7* 90 ± 17* 93 ± 11* <10 For each compound the drug concentration that showed the greatest mean inhibition is shown. NDGA, MK 886, and MK 591 (5 µM), AA861, Zileutin (10 µM ), ASA (100 µM ). All values were determined by assessment of % growth inhibition calculated from the optical density value, with a minimum of 6 replicates from at least three different experiments per cell line. *Percent growth inhibition compared to vehicle control was statistically significant (P<0.05).

Table 3 Cell Cycle Distribution _____________________________________________________________________________ % of Cells _________________________________________ Cell Line Treatment G1 S G2/M _________________________________________ MB-231 Untreated 41.9 ± 5 18.9 ± 0.4 38.5 ± 5.2 NDGA 2 µM 65 ± 8.3 10 ± 1.2 25.0 ± 9.5 NDGA 4 µM 77.8 ± 3.3 * 7.3 ± 2.8 * 14.8 ± 0.7* MK886 1 µM 63.5 ± 9.5 13.1 ± 1.8* 23.2 ± 7.9 MK886 2 µM 75 ± 0.3* 9.8 ± 1.6* 18.3 ± 1.2* MCF-7 Untreated 50.3 ± 8.7 15.8 ±1.1 33.5 ± 6.6 NDGA 2µM 49.4 ± 9.3 13.8 ± 1.0 37.1 ± 9.9 NDGA 4µM 61 ± 1.9 9.6 ± 0.1 30.1 ± 0.9 MK886 1 µM 64 ± 7.1 11.8 ± 0.7 24.7 ± 7.1 MK886 2 µM 63.7 ± 15.6 14 ± 18 22.3 ± 3.9 ____________________________________________________________________________ Exponentially growing cells were treated with NDGA and MK 886 for 24 hours, and cell cycle analysis was performed by flow cytometry. The data shown are the average ± s.d. of 2 experiments. *Statistical significance (P<0.05)

Fig. 1

Figure 1. Production of endogenous 5-HETE after exposure to two growth factors as detected by RIA. The cell lines used were MCF-7 (solid bars), ZR-75 (striped bars), T47D (checkered bars), and SKBR3 (open bars). The cells were homogenized, extracted onto C-18 disposable cartridges, and eluted using 85% acetonitrile/15% MeOH. Results are presented as pmol/0.1 ml of 5-HETE and its derivatives released after exposure to 10 µM/ml TF, and 5 µM/ml IGF-1. Error bars indicate SE. All 5-HETE values for the four cell lines were significantly increased from baseline values (P<0.05).

Fig. 2

5-H

PE

TE

5-H

ET

E

LTC

4

LTC

4 p

rod

s

LTB

4

LTB

4 p

rod

s

12-H

ET

E

15-H

PE

TE

15-H

ET

E

Lip

oxyn

TX

B2A

PG

's

Figure 2. HPLC characterization of AA metabolism in MCF-7 cells exposed to 2 ng/ml IGF-1 (open bars) and IGF-1 with the addition of 5 µM MK 886 (solid bars). The cells were incubated overnight with 100 µCI [3H] AA before extraction as outlined in Materials and Methods. The various metabolites were separated by RP-HPLC and quantitated using an online radioactivity detector. The values shown are major measurable products released and represented as fold increases over control. The data shown were obtained at 90 s, a time period showing maximal production of many metabolites of interest. The data represent the average of at least three experiments, and the SE of the values is shown as error bars. Statistical analysis was performed comparing fold change induced by IGF-1. The differences between cells exposed to MK 886 compared with control were statistically significant (P<0.05) for all of the metabolites shown except for 12-HETE. NS, not statistically significant.

Fig. 3

Figure 3. To show specificity of growth inhibition, we incubated breast cancer cell line ZR-75 for 12 h with MK 591 at various concentrations, using the in vitro proliferation assay as described in Materials and Methods. 5(S)-HETE was then added at nM concentrations, and the experiment was terminated after an additional 24 h. Open circles represent MK 591 alone, closed circles represent 15 nM 5(S)-HETE, open triangles represent 150 nM 5(S)-HETE, and closed triangles represent 1500 nM 5(S)-HETE.

Fig. 4

Figure 4. In vitro apoptosis analysis 24 h after exposure to two inhibitors. Control (open bars), treatment with 4 µM NDGA (striped bars), and 2 µM MK 886 (filled bars). Results are presented as percent apoptotic cells in five breast cancer cell lines after counting 100 cells by two independent readers. Experiments were conducted twice for each cell line. The error bars indicate standard deviation. All values were statistically different from control (P<0.05).

Fig. 5

Figure 5. Representative cytomorphological analysis for apoptosis. Cell line MCF-7 (A–C) and MB-231 (D–F) after exposure for 24 h to media control (A, D), 2 µM MK 886 (B, E), and 4 µM NDGA (C, F). This analysis was performed by DAPI staining analyzed via fluoresence microscopy. Cell line MB-231 (G–I) exposed to the same conditions as above (control, MK 886, NDGA, respectively) and analyzed using the TUNEL assay. (Magnification: A–F, ×2500; G–I, ×1600).

Fig. 6

Figure 6. A–D) Immunocytochemical analysis of cell line MCF-7 with antibodies against Bcl-2 (A, B) and Bax (C, D). A and C are untreated control cells; B and D were treated for 24 h with 2 µM MK 886, cytospins were performed, and the specimens were analyzed using a Zymed Histostain kit (A–D, magnification ×1600). Western blot analysis of the effect of LO inhibitors on levels of Bcl-2 and Bax protein (E). MCF-7 cells were treated with either 4 µM NDGA or 2 µM MK 886 for the time points indicated. Cell lysates were prepared, and equal amounts of protein were subjected to electrophoresis on 12% SDS-polyacrylamide gels and transferred to nitrocellulose membranes. Western blot analysis was performed using anti-Bcl-2 or anti-Bax antibodies using chemoluminescent detection.

Fig. 7

Figure 7. Effect of administration of 5-LO inhibitors on apoptosis (A) and tumor growth reduction (B) in vivo. Nude mice were injected with (107) MCF-7 cells, and palpable tumors were formed after 1 wk. The mice were administered daily NDGA (p.o), MK 591 (s.c), or PBS (s.c) for days 7–28, and tumor volume was measured biweekly. At termination of the study, tumors were harvested and histological sections of the flank xenografts were analyzed for evidence of apoptosis using the TUNEL assay (A). Data represent the average ± SD of 10 microscopic fields counted. The difference in the number of apoptotic cells between the treated samples and controls was statistically significant for both drugs (*P<0.05). In vivo growth reduction (B): Tumor volume was measured biweekly, the mean value ± SD of five determinations are indicated for each group (*statistical significance, P<0.05).

Fig. 8

Figure 8. Effect of 5-LO inhibitors on PPAR expression status (A–D), and effect of increased PPAR activation on in vitro breast cancer cell line growth (E, F). Breast cancer cell line ZR-75 was grown in the presence of 5 µM inhibitors for 24 and 48 h, total RNA was isolated, and RT-PCR was performed for PPARα and PPARγ using specific primers. Results are presented as percent of control after normalizing with the actin values. The experiment was repeated three times. The error bars indicate standard deviation. The first row (A, B) shows the effect of MK 886 on PPARα (A) and PPARγ (B) at baseline, 24 and 48 h. The second row shows the effect of NDGA on PPARα (C) and PPARγ (D) under the same conditions. The third row shows the growth inhibitory effect of PPAR ligands on breast cancer cell line ZR-75 (E) and T47D (F), as evaluated using a proliferation assay. All values were determined by assessment of percent growth inhibition calculated from the optical density value, with a minimum of six replicates from at least three different experiments per cell line. The ligands included WY-14643 (open circles), LY 171883 (closed circles), fenofibrate (triangles), and clofibrate (diamonds). The error bars indicate standard deviation. *The values were significantly different from control (P<0.05).

Fig. 9

Figure 9. Factors contributing to breast cancer growth inhibition or apoptosis after exposure to 5-LO inhibitors. A) Northern blot analysis for PPARγ (1.8 Kb) of breast cancer cell line T47D treated with 5 µM MK 886 for 6, 12, and 24 h and untreated control. RNA was isolated as described in Materials and Methods, and 10 µM of total RNA was loaded per lane. Ethidium bromide staining of 28S rRNA was used to check for equal loading and RNA integrity (lower panel). B) RNA isolated from T47D cells treated with 5 µM MK 886 for different time periods was subjected to RT-PCR using specific primers for PPARγ. The PCR products were normalized to actin gene and 18S. PCR products were quantified by densitometry, and the data points presented are expressed as percent of control and represent mean and SE of three experiments. C) Proliferation analysis of breast cancer cells, T47D. All cells were incubated in the indicated concentrations of either LY171883 (stars) or MK886 (open and closed circles) for 12 h. The drug was removed, the cells were refeed with media and incubated for an additional 24 h, and the experiments were terminated. All values were determined by assessment of percent growth inhibition calculated from the optical density value, with a minimum of six replicates from at least three different experiments. The error bars indicate standard deviation. D) Parallel experiments were performed as described for C to evaluate the effects on apoptosis, as reflected by the early apoptosis marker M30 (Roche Molecular Biochemicals, Indianapolis, IN). Open bars represent apoptosis with treatment of MK 886 alone and filled bars represent sequential treatment with MK 886 and 4 µM LY 171883. Results are presented as percent apoptotic cells for two concentrations of MK 886.